Macromolecules 2006, 39, 3327-3336

3327

Comparison of Nanocomposites Prepared from Sodium, Zinc, and

Lithium Ionomers of Ethylene/Methacrylic Acid Copolymers
Rhutesh K. Shah and D. R. Paul*
Department of Chemical Engineering and Texas Materials Institute, The UniVersity of Texas at Austin,
Austin, Texas 78712
ReceiVed January 2, 2006; ReVised Manuscript ReceiVed March 3, 2006

ABSTRACT: Morphology and mechanical properties of nanocomposites prepared by melt mixing a montmorillonite-based organoclay with lithium, sodium, and zinc ionomers of poly(ethylene-co-methacrylic acid) are
presented here. The effect of the type of neutralizing cation on the melt rheology and organoclay exfoliation
efficiency of the ionomers was examined using transmission electron microscopy, wide-angle X-ray scattering,
DSM axial-force rheometry, and stress-strain analysis. Nanocomposites prepared from the sodium and the zinc
ionomers displayed much better exfoliation of the organoclay and superior levels of reinforcement compared to
equivalent nanocomposites prepared from the lithium ionomer. On the basis of the detailed experimental analysis,
the lower levels of organoclay exfoliation observed in the nanocomposites prepared from the lithium ionomer are
suggested to be a result of irreversible exchange of quaternary ammonium ions (of the organoclay) for the very
small lithium ions (of the polymer) that can enter the montmorillonite lattice structure.

Introduction
A commercially important class of ionomers consists of
random copolymers of ethylene and methacrylic acid where
some of the acid groups are neutralized to form metal salts.1
Typically, the methacrylic acid content of these polymers is low
(<15 mol %), and the degree of neutralization ranges from 20
to 80%. The resulting polymer structure has three regions:
amorphous and crystalline phases plus ionic clusters.2 The
incorporation of the ionic groups improves the toughness, melt
viscosity, clarity, and adhesion properties of the copolymer.3
As a result, these ionomers are extensively used in a variety of
applications ranging from orthotics and prosthetics to packaging
and sealants and from glass coatings to outer coverings for golf
balls and bowling pins.4
Recently, nanocomposites prepared by melt mixing such
ionomers with organically modified montmorillonite clays
(organoclays) have attracted a great deal of technological and
scientific interest because of their potential for significant
improvements in physical and barrier properties at low filler
levels.5-7 The key to achieving these benefits is exfoliating the
organoclay into the polymer matrix to generate high aspect ratio
particles. In this regard, ethylene/methacrylic acid ionomers have
a distinct advantage over the base polyolefin, low-density
polyethylene (LDPE). The presence of the pendant ionic groups
and the polar methacrylic acid groups in these ionomers
potentially creates favorable interactions between the polymer
and the aluminosilicate clays, resulting in a much more
exfoliated morphology compared to nanocomposites prepared
from LDPE.7 Nanocomposites prepared from ionomers of
polypropylene,8 poly(ethylene terephthalate),9,10 poly(butylene
terephthalate),10,11 and a variety of other nonpolar thermoplastics
polymers12-14 also reveal good levels of organoclay exfoliation.
In this study, we examine the effect of the type of neutralizing
cation on the exfoliation efficiency of the ion-containing polymer
and on the morphology and mechanical properties of the
nanocomposites formed. Sodium, zinc, and lithium ionomers
* Corresponding author: Tel 512-471-5392; Fax 512-471-0542; e-mail
drp@che.utexas.edu.

of poly(ethylene-co-methacrylic) acid were carefully chosen

such that all other specifications, viz., melt index, acid content,
and degree of neutralization, are comparable. Nanocomposites
were prepared by melt mixing these ionomers with an appropriate organoclay in a twin-screw extruder. Stress-strain analysis,
X-ray scattering (WAXS), and transmission electron microscopy
(TEM) coupled with particle analysis were used to evaluate the
level of organoclay exfoliation, and appropriate mechanisms
have been suggested to explain the differences in the rheology,
morphology, and properties of these nanocomposites.
Experimental Section
Materials. Three commercial grades of Surlyn ionomer resins,
Surlyn 7940, 8945, and 9945, were purchased from du Pont. These
are lithium, sodium, and zinc salts, respectively, of poly(ethyleneco-methacrylic acid). As shown in Table 1, all three polymers have
similar melt indices, acid contents, and degrees of neutralization.
The organically modified clay, designated here as M2(HT)2, was
generously donated by Southern Clay Products and was used as
received. It was prepared by a cation exchange reaction between
sodium montmorillonite (Na-MMT) and a two-tailed quaternary
ammonium surfactant, dimethyl bis(hydrogenated-tallow) ammonium choloride (Arquad 2HT-75). Selected properties of this
organoclay are also included in Table 1. The choice of the
organoclay was based on a recent study exploring the effect of
surfactant structure on organoclay exfoliation in sodium ionomers
of poly(ethylene-co-methacrylic acid).15 The study revealed that
higher levels of organoclay exfoliation could be achieved using
surfactants with multiple alkyl tails on the ammonium ion rather
than one tail. This is believed to be the result of the better affinity
these ionomers have for the largely aliphatic organic modifier than
for the pristine surface of the clay. In addition, the larger the number
of alkyl tails, the more the silicate surface is shielded from the
matrix; this combination of factors leads to better exfoliation of an
organoclay having multiple alkyl tails.
Melt Processing. Melt compounded composites were prepared
using a Haake, corotating, intermeshing twin-screw extruder
(diameter ) 30 mm, L/D ) 10) using a barrel temperature of 190
C and a screw speed of 280 rpm. The polymers and the organoclays
were premixed and fed to the extruder using a single hopper. As
in the past,15 the feed rate was initially set at 1200 g/h. However,
the high melt viscosities generated in the case of a few nanocom-

10.1021/ma0600052 CCC: $33.50 2006 American Chemical Society

Published on Web 03/31/2006

3328

Shah and Paul

Macromolecules, Vol. 39, No. 9, 2006

Table 1. Materials Used in This Study

material

commercial
designation

specifications

supplier

zinc ionomer of
poly(ethylene-co-methacrylic acid)

Surlyn 9945

MI ) 4.1 g/10 min

specific gravity ) 0.97
methacrylic acid content ) 15.0 wt %
zinc oxide content ) 3.70 wt %
neutralization ) 40%

E.I. du Pont de
Nemours and Company

sodium ionomer of
poly(ethylene-co-methacrylic acid)

Surlyn 8945

MI ) 4.5 g/10 min

specific gravity ) 0.96
methacrylic acid content ) 15.2
sodium content ) 1.99 wt %
neutralization ) 40%

E.I. du Pont de
Nemours and Company

lithium ionomer of
poly(ethylene-co-methacrylic acid)

Surlyn 7940

MI ) 2.66 g/10 min

specific gravity ) 0.94
methacrylic acid content ) 14.7 wt %
lithium content ) 0.52 wt %
neutralization ) 40%

E.I. du Pont de
Nemours and Company

organoclay:a dimethyl
bis(hydrogenated-tallow)
ammonium montmorillonite

Cloisite 20A

organic loading ) 95 mequiv/100 g clay

organic content ) 39.6 wt %
d001 spacingb ) 25.5

Southern Clay Products

a The selected organoclay is designated as M (HT) in this study, where M ) methyl and HT ) hydrogenated tallow. Tallow is a natural product composed
2
2
predominantly (63%) of saturated and unsaturated C18 chains. HT is the saturated form yet still contains a small fraction of double bonds. b The basal
spacing corresponds to the characteristic Bragg reflection peak d001 obtained from a powder WAXS scan of the organoclay.

posites based on the zinc ionomer (containing more than 5 wt %

MMT) resulted in high values of extruder torque that exceeded the
permissible limits of the equipment. Hence, all nanocomposites
evaluated in this study were prepared using a lower feed rate of
800 g/h. Surlyn materials were dried in a vacuum oven at 65 C
for a minimum of 48 h prior to compounding while the organoclays
were used as received. In prior studies from this laboratory,16,17
the amount of montmorillonite in the nanocomposite was determined by placing predried nanocomposite pellets in a furnace at
900 C for 45 min and weighing the remaining MMT ash. It was
not possible to employ this technique with these ionomers since
the polymer itself resulted in a hard, yellowish-green coating on
the inside of the crucible reflecting some complex residue of the
inorganic component of the ionomer. The amount of the residue
varied from batch to batch, rendering this method useless for
quantitative analysis. Hence, to ensure that a predetermined
polymer/MMT ratio was maintained in all cases, the desired
amounts of clay and polymer were premixed before feeding to the
extruder, and precautions were taken to minimize any losses of
organoclay during the extrusion process.
Tensile specimens (ASTM D638) were prepared by injection
molding using an Arburg Allrounder 305-210-700 injection molding
machine using a barrel temperature of 220 C, mold temperature
of 45 C, injection pressure of 70 bar, and a holding pressure of
40 bar. After molding, the samples were immediately sealed in a
polyethylene bag and placed in a vacuum desiccator for a minimum
of 24 h prior to testing. The samples made from the three unfilled
ionomers were colorless and clear. On the other hand, the injectionmolded nanocomposite samples had a yellowish tint, the intensity
of which increased with the clay content. There were visible
differences between the clarity of the nanocomposites prepared from
the three ionomers. Nanocomposite samples prepared from the
sodium and the zinc ionomers maintained the clarity of the unfilled
polymer samples. In contrast, the clarity of the samples prepared
from the lithium ionomer decreased as the clay content increased;
this reflects the poorer exfoliation of organoclay in this matrix.
Testing and Characterization. Tensile tests were conducted at
room temperature according to ASTM D696 using an Instron model
1137 machine equipped with digital data acquisition capabilities.
Modulus was measured using an extensiometer at a crosshead speed
of 0.51 cm/min. Elongation at break and tensile strength at break
were measured at a crosshead speed of 5.1 cm/min. Typically, data
from six specimens were averaged to determine the tensile
properties with standard deviations of the order of 2-7% for
modulus, 1-10% for tensile strength at break and, 2-21% for

elongation at break. Yield stress data are not reported because the
stress-strain behavior of some of the nanocomposites do not show
a distinct yield point.
Relative melt viscosities of the polymers and their nanocomposites were determined using a DSM microcompounder (chamber
volume ) 5 cm3) operated under recycle mode. Prior to testing,
the polymers and their nanocomposites were dried in a vacuum
oven at 65 C for 48 h. A charge of 2.5 g of polymer/
nanocomposite pellets was mixed separately in the microcompounder using a constant screw speed of 100 rpm at 190 C. The
axial force, which is a function of the melt viscosity, was recorded
at regular intervals over a 10 min period.
WAXD was conducted using a Sintag XDS 2000 diffractometer
in the reflection mode with an incident X-ray wavelength of 1.542
at a scan rate of 1.0/min. X-ray analysis was performed at room
temperature on injection-molded Izod bars. The specimens were
oriented such that the incident beam reflected off the major face.
Samples for TEM analysis were taken from the core portion of
an Izod bar parallel to the flow direction but perpendicular to the
major face. Ultrathin sections 50 nm in thickness were cut with
a diamond knife at a temperature of -65 C using a Reichert-Jung
Ultracut E microtome. Sections were collected on 300 mesh grids
and subsequently dried with filter paper. These were then examined
using a JEOL 2010F TEM equipped with a field emission gun at
an accelerating voltage of 120 kV at a magnification of 25K. The
negative films containing the electron micrographs were electronically scanned and converted into gray scale tagged-image file format
(TIFF) files, and the images were subsequently printed on a 20.3
25.4 cm2 photographic paper. To conduct quantitative analysis
on these images, the lengths, thicknesses, and aspect ratios of the
particles were determined. The lengths of the particles were traced
manually on a transparency film using a felt tip pen, and the
resultant tracings were scanned and converted into TIFF files. The
thickness of the dispersed platelets and agglomerates were traced
digitally on an overlapped blank layer in Adobe Photoshop under
high magnification. Thus, two separate tracings were done for each
TEM picture: one contained the lengths of the particles, and the
other one contained their thicknesses. The resulting black and white
layer files were then imported into an image analysis software, Scion
Image, which analyzed the traced particles, assigned a numerical
label to each of them, and exported their characteristic dimensions
to a different file. Since two different tracings were used for
measuring the lengths and thicknesses of the particles, each particle
got assigned two different numerical labels. This made it extremely
difficult to match the length of a given particle with its thickness

Macromolecules, Vol. 39, No. 9, 2006

Morphology of Nanocomposites 3329

Figure 1. Electrical current drawn by the extruder motor during

extrusion of the various nanocomposites based on ionomers of ethylene/
methacrylic acid copolymers. The feed (polymer + organoclay) rate
in all cases is 800 g/h.

and, thus, calculate the aspect ratio of a particular particle. Hence,

in this study, the particle aspect ratio for any given nanocomposite
was determined by dividing the average particle length by the
average particle thickness.18-20

Observations and Results

Processability and Rheology. Figure 1 shows the electrical
current drawn by the extruder motor while processing the
various nanocomposites based on lithium, sodium, and zinc
ionomers of poly(ethylene-co-methacrylic acid) at a feed rate
of 800 g/h. At a fixed screw speed, the current drawn is a
function of the torque, which in turn is dependent upon the melt
viscosity of the nanocomposite being processed. For nanocomposites prepared from the lithium ionomer, the current does not
change much with clay concentration. For the nanocomposites
based on the sodium ionomer, there is a small increase in the
current when the clay content is gradually increased from 2.5
wt % MMT to 10.0 wt % MMT. The slightly lower current
observed for the nanocomposites made from the sodium
ionomer, compared to those formed from the lithium ionomer
at low filler concentrations, could be a result of the small
difference in the melt indices of the two matrices (4.5 g/10 min
vs 2.6 g/10 min). In contrast, the current for the nanocomposites
based on the zinc ionomer increases steadily, but significantly,
when the montmorillonite content is increased from 2.5 to 10
wt %, which is indicative of a large increase in melt viscosity
as the clay content is increased.
To get a rough idea of the comparative melt viscosities, we
decided to measure the relative viscosities of the unfilled
polymers and their nanocomposites using a DSM microcompounder, as described in the Experimental Section. Figure 2a
compares the axial force generated by the polymer melts when
sheared at 190 C using a screw speed of 100 rpm in a DSM
microcompounder. Initially (t < 90 s), the forces decrease with
time for all three ionomers. The differences between the forces
generated by the three polymer melts in this region could be
attributed to the small differences in their melt indices. Between
t ) 90 s and t ) 600 s, there is little change in the axial force
generated by the lithium and sodium ionomer melts; however,
the force exerted by the zinc ionomer melt increases steadily to
a value which is 25% higher than the minimum observed at
t < 90 s. Although we are not completely sure what causes this
increase in the melt viscosity of the zinc ionomers, we believe
it is a result of possible formation of anhydrides in the polymer

Figure 2. Axial forces generated by (a) the ionomers and (b)

nanocomposites containing 10 wt % MMT prepared from these
ionomers, when sheared at 190 C using a screw speed of 100 rpm in
a DSM microcompounder.

melt. As mentioned in Table 1, the degree of neutralization of

the ionomers used in this study is about 40%. The remaining
unneutralized acid groups are capable of forming anhydrides
(with the expulsion of water); these anhydrides could act as
branch (and eventually cross-link) points, which could lead to
an increase in the melt viscosity of the polymer. Transition metal
salts (such as zinc acetate) are well-known to catalyze anhydride
formation, just as they act as esterification catalysts. So, it could
be that this process is more rapid for Zn ionomers than Li and
Na ionomers.21,22 Small amounts of water in the polymer could
affect the anhydride formation reaction. In this study, the pellets
were dried in a vacuum oven prior to testing, as described in
the Experimental Section; however, no additional precautions
were taken to prevent the absorption of moisture during the
weighing and charging of polymers to the microcompounder.
Also, no special steps were taken to prevent or facilitate the
removal of water vapor from the partially filled chamber during
the melt mixing process.
The nanocomposites prepared from these ionomers exhibited
similar trends as shown in Figure 2b. Careful comparison of
parts a and b of Figure 2 suggests that, while the axial forces
generated by the nanocomposites prepared from sodium and
zinc ionomers are higher than those of the corresponding unfilled
polymers, the nanocomposite based on the lithium ionomer
produces a smaller viscous force than the lithium ionomer itself.
The higher melt viscosities of the sodium and zinc ionomer
nanocomposites could be an artifact of organoclay exfoliation
in these polymers, while the lower melt viscosity of the Li
ionomer nanocomposites could be a consequence of composi-

3330

Shah and Paul

Macromolecules, Vol. 39, No. 9, 2006

Figure 3. TEM micrographs of nanocomposites prepared from M2(HT)2 organoclay and (a) lithium, (b) sodium, and (c) zinc ionomers of poly(ethylene-co-methacrylic acid). The concentration of MMT in all three cases is 5 wt %. Sections were microtomed from the core portion of an
Izod bar in a plane parallel to the flow direction but perpendicular to the major face.

tional changes of the organoclay that are unique to mixing with

the lithium ionomer as described later.
The results of this limited rheological evaluation of the
polymers and their nanocomposites suggest a more comprehensive rheological study; however, this is beyond the scope
of this paper.
TEM and Particle Analysis. Figure 3 shows TEM micrographs comparing the morphology of nanocomposites formed
from M2(HT)2 organoclay and lithium, sodium, and zinc
ionomers of poly(ethylene-co-methacrylic acid). The concentration of montmorillonite in all three samples is 5 wt %.
Nanocomposites prepared from the sodium ionomer (Figure 3b)
and the zinc ionomer (Figure 3c) exhibit better clay exfoliation
or dispersion of the organoclay than the one prepared from the
lithium ionomer (Figure 3a). The micrographs of nanocomposites prepared from the former two reveal a pattern of uniformly
dispersed single platelets along with a few thin stacks comprised
of 2-4 platelets. On the other hand, the morphology of the
lithium ionomer based nanocomposites revealed a significant
number of thicker stacks of platelets.
To provide a quantitative comparison of the level of organoclay exfoliation in the three matrices, particle analysis was
conducted on TEM micrographs of nanocomposites using image
analysis techniques as described in the Experimental Section
and in previous studies.18-20,23 Figure 4 shows a series of
histograms of MMT particle lengths and pertinent statistical data
obtained on nanocomposites containing 5 wt % MMT
prepared from M2(HT)2 organoclay and the three ionomers. The
sections were taken parallel to the flow direction but perpendicular to the major face. Similar measurements were conducted

for the thickness of clay particles, and the results are plotted in
Figure 5. The filler particles in the nanocomposites prepared
from the sodium and the zinc ionomer were shorter and thinner
than the ones in the nanocomposites based on the lithium
ionomer. The average particle length of 92 nm calculated in
the case of the former two nanocomposites agrees well with
the average particle length in well-exfoliated nylon-6 nanocomposites determined using a similar technique.18 It is interesting
to note the greater average particle length in the nanocomposite
prepared from the lithium ionomer (163 nm) compared to similar
nanocomposites prepared from the sodium and zinc ionomers
(92 nm). This could be a result of a partial breakdown of the
clay agglomerates comprised of randomly overlaid platelets and
the skewing of thicker clay bundles as described by Chavarria
et al.19 The thickness distribution profiles of particles in the
nanocomposites prepared from the sodium ionomer and the zinc
ionomer appear to be fairly similar, with 75% of the particles
having a thickness of 4 nm or less. Their average particle
thickness (3.5 nm) was noticeably lower than that calculated
for the nanocomposites based on the lithium ionomer (11.9 nm).
The aspect ratio of the particles in each nanocomposite was
determined by dividing the average particle length by the
average particle thickness of the nanocomposite (see Table 2).
The average particle aspect ratio in the nanocomposites based
on the sodium ionomer and the zinc ionomer was calculated to
be 27, while that in the nanocomposites based on the lithium
ionomer was determined to be 14. On the basis of the TEM
evaluation and particle analysis, it can be concluded that (i) the
sodium and the zinc ionomers are much more efficient at
exfoliating the M2(HT)2 organoclay than the lithium ionomer

Macromolecules, Vol. 39, No. 9, 2006

Morphology of Nanocomposites 3331

Figure 4. Histograms of MMT particle length obtained by analyzing

TEM micrographs of nanocomposites containing 5 wt % MMT
prepared from (a) lithium, (b) sodium, and (c) zinc ionomers of poly(ethylene-co-methacrylic acid).

Figure 5. Histograms of MMT particle thickness obtained by analyzing

TEM micrographs of nanocomposites containing 5 wt % MMT
prepared from (a) lithium, (b) sodium, and (c) zinc ionomers of poly(ethylene-co-methacrylic acid).

and (ii) there is not much difference in the morphology of the

nanocomposites (containing 5 wt % MMT) prepared from the
sodium and the zinc ionomer.
Mechanical Properties. Selected mechanical properties of
the nanocomposites prepared are listed in Table 3. However,
before we discuss these properties in detail, it is important to

highlight the similarities and subtle differences in the stressstrain behavior of these composites. Figure 6 displays typical
stress-strain diagrams for nanocomposites prepared from M2(HT)2 organoclay and lithium, sodium, and zinc ionomers. The
stress-strain curves of the unfilled lithium and sodium ionomers
reveal a distinct stress maximum followed by a slight drop in

3332

Shah and Paul

Macromolecules, Vol. 39, No. 9, 2006

Table 2. Particle Analysis Results

organoclay
nanocomposites

total no. of
particles

number-average
particle length
(nm)

weight-average
particle length
(nm)

number-average
particle thickness
(nm)

weight-average
particle thickness
(nm)

aspect
ratioa

aspect
ratiob

lithium ionomer (Surlyn

7940) + 5 wt % MMT
sodium ionomer (Surlyn
8945) + 5 wt % MMT
zinc ionomer (Surlyn
9945) + 5 wt % MMT

502

163.3

214.7

11.9

25.4

13.7

8.4

762

92.6

119.2

3.5

8.0

26.5

14.9

703

92.7

120.6

3.4

5.5

27.3

21.9

a These values of the aspect ratio were computed from the number-average platelet lengths and thicknesses. b These values of the aspect ratio were
computed from the weight-average platelet lengths and thicknesses.

Table 3. Selected Mechanical Properties of Nanocomposites Prepared by Melt Mixing M2(HT)2 Organoclay and Ionomers of
Poly(ethylene-co-methacrylic acid)
polymer

montmorillonite
content (wt %)

tensile modulus,
E (GPa)

relative modulus,
E/Em

elongation at break
(5.1 cm/min) (%)

tensile strength at break

(5.1 cm/min) (MPa)

zinc ionomer
zinc ionomer
zinc ionomer
zinc ionomer
sodium ionomer
sodium ionomer
sodium ionomer
sodium ionomer
lithium ionomer
lithium ionomer
lithium ionomer
lithium ionomer

0.0
2.5
5.0
10.0
0.0
2.5
5.0
10.0
0.0
2.5
5.0
10.0

0.176
0.314
0.447
0.795
0.260
0.412
0.568
0.908
0.292
0.407
0.491
0.676

1.00
1.78
2.50
4.51
1.00
1.58
2.18
3.49
1.00
1.38
1.68
2.32

172
116
85.9
58.9
194
130
119
65.6
136
116
104
98.4

19.3
22.2
24.2
29.2
21.0
22.6
25.9
28.8
21.0
24.2
24.8
27.2

the tensile stress after this yield point, corresponding to the onset
of necking. This stress drop gradually diminishes as the clay
content increases. In contrast, the stress-strain curves of the
zinc ionomer and nanocomposites prepared from it do not show
a maximum stress or a yield point. The stress-strain curves of
almost all polymers/nanocomposites suggest the occurrence of
strain hardening. The slope of the plastic region, indicative of
the level of strain hardening, decreases as the clay content
increases.
Figure 7a compares the tensile moduli of the different
nanocomposites as a function of their montmorillonite content.
As expected, the stiffness of all ionomers improves substantially
with addition of M2(HT)2 organoclay. To get a comparative idea
of the improvements in the level of reinforcement achieved,
the relative moduli of the nanocomposites are plotted against
the montmorillonite content in Figure 7b. It is clear that the
increase in modulus observed in nanocomposites prepared from
the sodium and zinc ionomers is much greater than that observed
in nanocomposites prepared from the lithium ionomer. However,
the level of reinforcement observed in nanocomposites formed
from the zinc ionomer seems to be somewhat greater than that
seen for those formed from the sodium ionomer. These
differences are more pronounced at higher MMT content. Thus,
it appears that the moduli data are not in complete agreement
with the results of the particle analysis, which revealed similar
aspect ratios for the zinc ionomer and sodium ionomer-based
nanocomposites containing 5 wt % MMT. We are not completely sure about the cause of this disagreement between the
results of the two analytical methods. In principle, some part
of this discrepancy could stem from differences between the
moduli of the two ionomers (0.176 GPa vs 0.260 GPa).
Composite theory predicts that for a given filler aspect ratio
low-modulus matrices offer greater potential for reinforcement
per unit mass of filler than high-modulus matrices due to the
larger ratio of filler modulus to the matrix modulus.24,25
However, simple calculations using the Halpin-Tsai theory
reveal that the discrepancy noted above cannot be fully
rationalized in this way.26,27 But then again, these calculations

are based upon numerous simplifications and assumptions of

ideal conditions.26
Another explanation may stem from some departure from one
of the basic premises of composite theory; i.e., the matrix
modulus is not altered by the presence of the nanoscale clay
particles.24,25 It was suggested above that anhydride formation
occurs in nanocomposite based on the zinc ionomer but not those
containing sodium or lithium. This chemistry might possibly
alter the material properties as clay is added. There seems to
be no simple or straightforward way to test such a hypothesis.
Finally, a certain degree of caution should be exercised when
attempting to correlate mechanical properties with morphology
determined by TEM since each micrograph provides a snapshot
of only a small microscopic area. An extensive particle analysis
is needed to fully describe the morphology of the macroscopic
sample which is reflected in property measurements like
modulus.
The comparatively smaller improvements in the modulus of
the lithium ionomer by addition of the organoclay is a result of
the relatively lower levels of organoclay exfoliation in these
systems compared to nanocomposites prepared from the zinc
and lithium ionomers (as revealed by TEM micrographs). The
slightly higher modulus of the unfilled lithium ionomer compared to the other two polymers contributes to some degree to
this observation; we propose an additional explanation below.
The relationship between the MMT content of the nanocomposites and their elongation at break is shown in Figure 8. The
more exfoliated and stiffer nanocomposites prepared from the
zinc and sodium ionomers are less ductile than the relatively
less exfoliated nanocomposites prepared from the lithium
ionomer; generally, ductility decreases when stiffness is increased by reinforcement.7,15,28 Also, as mentioned in the
Introduction, the presence of the ionic clusters in these ionomers
improves their toughness compared to corresponding ethylene/
methacrylic acid copolymers. It is unclear whether the ductility
of the ionomers is lowered due to possible disruption of the
ionic aggregates by the exfoliated/partially exfoliated clay
particles.

Macromolecules, Vol. 39, No. 9, 2006

Morphology of Nanocomposites 3333

Figure 7. (a) Tensile modulus and (b) relative modulus of nanocomposites prepared from lithium, sodium, and zinc ionomers of poly(ethylene-co-methacrylic acid).

Figure 6. Representative stress-strain diagrams of nanocomposites

prepared from M2(HT)2 organoclay and (a) lithium, (b) sodium, and
(c) zinc ionomers of poly(ethylene-co-methacrylic acid).

WAXS Analysis. Figure 9 compares the WAXS scans of

the M2(HT)2 organoclay and its nanocomposites prepared by
melt mixing with the three ionomer matrices. The organoclay
pattern reveals an intense peak at around 2 ) 3.46, corresponding to a basal spacing of 25.5 . WAXS patterns of all
nanocomposites show a distinct peak indicative of the presence
of unexfoliated clay tactoids. However, the positions of the peaks
for the nanocomposites shift in different directions relative to
the peak of the organoclay. The peaks for composites formed
from the zinc and sodium ionomers are shifted to higher d
spacings than the organoclay, which according to prevalent

Figure 8. Elongation at break measured at a crosshead speed of 5.1

cm/min for nanocomposites prepared from lithium, sodium, and zinc
ionomers of poly(ethylene-co-methacrylic acid).

understanding suggests the intercalation of mass, e.g., polymer,

within the clay galleries. On the other hand, the peak for the
nanocomposite prepared from the lithium ionomer is shifted to
the right (lower d spacings), suggesting loss of mass from the
galleries, e.g., loss of surfactant mass by degradation or some
other mechanism.29 A detailed description and a proposed
mechanism for this phenomenon resulting from the interactions
between the lithium ionomer and the filler are described below.
The position of the peak did not change with organoclay
content of the nanocomposites; however, the height of the peak

3334

Shah and Paul

Macromolecules, Vol. 39, No. 9, 2006

Figure 9. WAXS patterns of nanocomposites prepared from M2(HT)2

organoclay and the three ionomers. The concentration of MMT in all
cases is 5 wt %. X-ray pattern of the M2(HT)2 organoclay is plotted
for comparison and is shifted vertically for clarity. The dotted vertical
line shows the position of the d001 peak of the organoclay.

increases as the clay concentration increases, as shown in Figure

10. WAXS results for nanocomposites prepared from unmodified polyethylene exhibit similar trends.23
Discussion
As described above, a series of polymer-silicate nanocomposites were prepared by melt mixing M2(HT)2 organoclay with
lithium, sodium, and zinc ionomers of poly(ethylene-co-methacrylic acid). Three observations merit special consideration:
(i) Unlike the nanocomposites prepared from the sodium and
the zinc ionomers, the melt viscosity of the nanocomposite
prepared from the lithium ionomer was lower than that of the
matrix polymer. (ii) Nanocomposites prepared from the zinc
and the sodium ionomers exhibited a higher level of organoclay
exfoliation compared to equivalent nanocomposites prepared
from the lithium ionomer. (iii) WAXS peaks of the nanocomposites prepared from the sodium and the zinc ionomers shifted
to the left (higher d spacing), whereas those based on the lithium
ionomer shifted to the right (lower d spacing).
The right shift of the WAXS peak of a nanocomposite is
indicative of a decrease in the interplatelet spacing of the
organoclay and is generally attributed to the loss of mass from
the organoclay galleries, e.g., by thermal degradation of the
surfactant.29,30 The quaternary ammonium surfactants used for
preparing the organoclays are known to degrade at the high
temperatures required to melt process the nanocomposites. The
initial thermal degradation, which is believed to follow a
Hoffman elimination mechanism,31-33 begins at temperatures
as low as 155-165 C. WAXS patterns of similar nanocomposites prepared from polyethylene,29 polypropylene,29 polystyrene,30 and nylon-6619 have also revealed such shifts to the
right. For LDPE/ M2(HT)2 nanocomposites, the X-ray peak
shifts to 23.2 when melt processed at 200 C.29 Increasing
the processing temperature to 240 C results in a further decrease
in the organoclay d spacing (22 ). For the nanocomposites
prepared from the lithium ionomer at 190 C, the organoclay d
spacing is reduced much more significantly (from 25.5 to 17.4
) compared to that reported for nanocomposites prepared from
other polymers using a similar organoclay.29 It is important to
note that WAXS patterns of nanocomposites prepared from the

Figure 10. WAXS patterns of nanocomposites prepared from M2(HT)2

organoclay and (a) lithium, (b) sodium, and (c) zinc ionomers of poly(ethylene-co-methacrylic acid). The dotted vertical line showing the
position of the d001 peak of the M2(HT)2 organoclay is included for
comparison. The curves are shifted vertically for clarity.

Macromolecules, Vol. 39, No. 9, 2006

Morphology of Nanocomposites 3335

Conclusion

Figure 11. Schematic of the proposed ion-exchange process between

the ionomer and the organoclay which subsequently leads to a reduction
in the d spacing of the nanocomposites prepared from the lithium
ionomer.

zinc and the sodium ionomers do not reveal such peak shifts.
As far as we know, there are no reports of the catalytic effects
of the Li+ ion on the Hoffman elimination reaction. Thus, it
appears that for the nanocomposites based on the lithium
ionomer there may be another mechanism than thermal degradation of the surfactant by the Hoffman elimination reaction
involved in the shift of WAXS peak to higher d spacings.
A possible explanation may involve an ion exchange process
between the organoclay and these ionomers, wherein a few of
the bulky quaternary ammonium ions of the organoclay are
replaced by the smaller metal cations of the ionomers, thus
reducing the d spacing of the organoclay as shown in Figure
11. The obvious question is, why is such behavior only exhibited
by the nanocomposites prepared from the lithium ionomer and
not by the nanocomposites prepared from the sodium and the
zinc ionomers? The answer possibly lies in the smaller size and
the higher reactivity of the lithium cation. The Li+ ion has a
radius of 0.68 compared to 0.74 for the Zn2+ ion and 0.98
for the Na+ ion. The Li+ ion is capable of entering the
montmorillonite lattice structure, resulting in an irreversible
exchange of quaternary ammonium ions.34-41 Hofmann and
Klemen34 showed that heating Li+-saturated bentonite caused
fixation of previously exchangeable Li+ ions and a reduction
of the cation exchange capacity (CEC) and expandability of
the main clay mineral present in bentonite (Hofmann-Klemen
effect). The irreversible exchange of the Li+ ions could serve
as a driving force for the cation exchange process between the
organoclay and the ionomer. On the other hand, Zn2+ and Na+
ions are not as capable of entering the montmorillonite lattice
structure as the Li+ ions. Emmerich et al.,42,43 while investigating
homoionic forms of montmorillonite heated at 220 C for 20 h,
found that the CEC of the Li+ forms of montmorillonite dropped
to 28% of the value observed in an unheated sample. In
comparison, the CEC of Zn2+ form dropped marginally to 91%,
and there was no change in the CEC of the Na+ form. Thus, in
the case of nanocomposites prepared from the sodium and the
zinc ionomers, there is no driving force to promote the cation
exchange reaction between the polymer and the organoclay.
The cation exchange process and the relatively lower levels
of organoclay exfoliation could also explain why nanocomposites prepared from the lithium ionomer do not show similar
improvements in melt viscosity over that of the unfilled ionomer
as those revealed by nanocomposites prepared from the sodium
and the zinc ionomers. Replacement of the smaller Li+ cations
by bulkier quaternary ammonium cations in the polymer phase
should affect the melt rheology of the nanocomposite.

Structure-property relationships for nanocomposites prepared

from ionomers of poly(ethylene-co-methacrylic acid) and M2(HT)2 organoclay have been presented here. The effect of the
neutralizing cation on the exfoliation efficiency of the polymer
was evaluated by comparing the morphology and properties of
nanocomposites prepared from sodium, zinc, and lithium
ionomers of poly(ethylene-co-methacrylic acid). On the basis
of transmission electron microscopy and particle analysis of the
images, nanocomposites prepared from the zinc and the sodium
ionomers show much better exfoliation of the organoclay
compared to equivalent nanocomposites prepared from the
lithium ionomer. The mechanical properties of the nanocomposites paralleled the TEM observations in this regard. WAXS
peaks of the sodium and the zinc ionomer based nanocomposites
showed shifts to larger d spacings relative to the organoclay,
suggesting intercalation of polymer species within the organoclay gallery. On the other hand, the WAXS peaks of nanocomposites prepared from the lithium ionomer shifted to lower d
spacings, indicating loss of mass from the galleries of the
organoclay. On the basis of WAXS analysis, mechanical
property data, and melt rheology observations, the lower levels
of organoclay exfoliation observed in the nanocomposites
prepared from the lithium ionomer are suggested to be a result
of the irreversible exchange of quaternary ammonium ions for
the very small lithium ions that can enter the montmorillonite
lattice structure.
Acknowledgment. The authors sincerely thank Doug Hunter
and Southern Clay Products for providing the organoclay and
WAXD analyses of the nanocomposites. We also acknowledge
Prof. R. A. Register of Princeton University and Prof. K. I.
Winey of University of Pennsylvania for providing useful
insights into rheological behavior of Surlyn ionomers.
References and Notes
(1) Eisenberg, A.; Kim, J.-S. Introduction to Ionomers; Wiley: New York,
1998.
(2) Technical Information, E. I. du Pont de Nemours and Company, http://
www.dupont.com/industrial-polymers/plastics/polymers/surlyn.html.
(3) Jerome, R.; Mazurek, M. In Ionomers, 1st ed.; Tant, M. R. M.,
Kenneth, A., Wilkes, G. L., Eds.; Blackie Academic and Professional: London, 1997; pp 3-40.
(4) Manufacturers information, E. I. du Pont de Nemours and Company,
http://www2.dupont.com/Surlyn/en_US/uses_apps/.
(5) Kobayashi, T.; Takahashi, T.; Monma, T.; Kurosaka, K.; Arai, T. WO,
9731973, 1997.
(6) Hasegawa, N.; Kawakado, M.; Usuki, A.; Okada, A. JP, 97-152854,
1998.
(7) Shah, R. K. Ph.D. Thesis, The University of Texas at Austin, 2006.
(8) Ding, R.-d.; Newell, C. US, 6,861,481, 2005.
(9) Barber, G. D.; Calhoun, B. H.; Moore, R. B. Polymer 2005, 46, 67066714.
(10) Barber, G. D.; Bellman, S. P.; Moore, R. B. Annu. Tech. Conf.sSoc.
Plast. Eng. 2003, 61st, 1369-1373.
(11) Chisholm, B. J.; Moore, R. B.; Barber, G.; Khouri, F.; Hempstead,
A.; Larsen, M.; Olson, E.; Kelley, J.; Balch, G.; Caraher, J.
Macromolecules 2002, 35, 5508-5516.
(12) Start, P. R.; Mauritz, K. A. J. Polym. Sci., Part B: Polym. Phys. 2003,
41, 1563-1571.
(13) Kovarova, L.; Kalendova, A.; Malac, J.; Vaculik, J.; Malac, Z.;
Simonik, J. Annu. Tech. Conf.sSoc. Plast. Eng. 2002, 60th, 22912295.
(14) Barber, G. D.; Carter, C. M.; Moore, R. B. Annu. Tech. Conf.sSoc.
Plast. Eng. 2000, 58th, 3763-3767.
(15) Shah, R. K.; Hunter, D. L.; Paul, D. R. Polymer 2005, 46, 26462662.
(16) Shah, R. K.; Paul, D. R. Polymer 2004, 45, 2991-3000.
(17) Fornes, T. D.; Yoon, P. J.; Keskkula, H.; Paul, D. R. Polymer 2001,
42, 9929-9940.
(18) Fornes, T. D.; Paul, D. R. Polymer 2003, 44, 4993-5013.
(19) Chavarria, F.; Paul, D. R. Polymer 2004, 45, 8501-8515.

3336

Shah and Paul

(20) Lee, H.-s.; Fasulo, P. D.; Rodgers, W. R.; Paul, D. R. Polymer 2005,
46, 11673-11689.
(21) Winey, K. I. Personal communication.
(22) Register, R. A. Personal communication.
(23) Shah, R. K.; Cui, L.; Williams, K. L.; Bernard, B.; Paul, D. R. J.
Appl. Polym. Sci., in press.
(24) Halpin, J. C.; Finlayson, K. M.; Ashton, J. E. Primer on Composite
Materials Analysis, 2nd rev./ed.; Technomic Pub. Co.: Lancaster, PA,
1992.
(25) Halpin, J. C.; Kardos, J. L. Polym. Eng. Sci. 1976, 16, 344-352.
(26) Fornes, T. D.; Paul, D. R. Macromolecules 2004, 37, 7698-7709.
(27) Stretz, H. A.; Paul, D. R.; Cassidy, P. E. Polymer 2005, 46, 38183830.
(28) Fornes, T. D.; Yoon, P. J.; Hunter, D. L.; Keskkula, H.; Paul, D. R.
Polymer 2002, 43, 5915-5933.
(29) Shah, R. K.; Paul, D. R. Polymer, in press.
(30) Tanoue, S.; Utracki, L. A.; Garcia-Rejon, A.; Tatibouet, J.; Cole, K.
C.; Kamal, M. R. Polym. Eng. Sci. 2004, 44, 1046-1060.
(31) Xie, W.; Gao, Z.; Pan, W.-P.; Hunter, D.; Singh, A.; Vaia, R. Chem.
Mater. 2001, 13, 2979-2990.
(32) Fornes, T. D.; Yoon, P. J.; Paul, D. R. Polymer 2003, 44, 75457556.

Macromolecules, Vol. 39, No. 9, 2006

(33) Davis, R.; Gilman, J.; VanderHart, D. Polym. Degrad. Stab. 2003,
79, 111-121.
(34) Hofmann, U.; Klemen, R. Z. Anorg. Chem. 1950, 262, 95-99.
(35) Madejova, J.; Bujdak, J.; Gates, W. P.; Komadel, P. Clay Miner. 1996,
31, 233-241.
(36) Madejova, J.; Bujdak, J.; Petit, S.; Komadel, P. Clay Miner. 2000,
35, 739-751.
(37) Alvero, R.; Alba, M. D.; Castro, M. A.; Trillo, J. M. J. Phys. Chem.
1994, 98, 7848-7853.
(38) Theng, B. K. G.; Hayashi, S.; Soma, M.; Seyama, H. Clays Clay Miner.
1997, 45, 718-723.
(39) Calvet, R.; Prost, R. Clays Clay Miner., Proc. Conf. 1971, 19, 187191.
(40) Stackhouse, S.; Coveney, P. V. J. Phys. Chem. B 2002, 106, 1247012477.
(41) Williams, J.; Purnell, J. H.; Ballantine, J. A. Catal. Lett. 1991, 9, 115119.
(42) Emmerich, K.; Madsen, F. T.; Kahr, G. Clays Clay Miner. 1999, 47,
591-604.
(43) Emmerich, K.; Plotze, M.; Kahr, G. Appl. Clay Sci. 2001, 19, 143-154.

MA0600052